retrospectiveworkbenchfandomcom-20200214-history
Tagisang 2014
Tagisang Robotics 2014 is the 4th iteration of the Tagisang Robotics: Design, Build, and Play. Here, we will document the journey of 2014's Tagisang season. Well it all starts with the Tagisang Workshop in May 2014. The only thing noteworthy was that the pneumatics and servos were introduced and we had a mini contest with that. Fast forward to Tagisang Kickoff, big changes to the rules which were in favor of more robot oriented gameplay and many changes to the track. GameMechanics FinalGuidelines The Plan We were planning to become a shooter again before the updated rules came out. Here is our most ambitious iteration of it at least. This time it would've been a maximized shooter with a ball dribbler long enough to gather 4 balls at a time. The launcher would be a long pvc pipe, that could have a twin pvc pipe to work together, covered with bed foam or rubber. The launcher would be powered by a ripped apart 18v cordless drill connected as like a chain drive or belt drive from the mouth of the drill to the axle of the pvc pipe. It would've had a top hopper as well on top of the robot to feed flying balls into the shooter as well. It would've sported ramps, that were either within the size limit or be extensions themselves, along its sides as a passive defense against robots that would try to push it. The drive train would've been a six wheeler design, with pocketbike motors drive the six smaller than usual wheels. The pocketbike motors are a leap forward compared to the usual wiper motors. It was faster and stronger on all levels compared to the wiper motors, perhaps only power comsumption was the problem. The idea is largely inspired by FRC 2008 Lunar Lunacy robots found in Youtube. This is the website of a local pocketbike seller: http://www.mariolan.com/spare-parts.html The Real Plan Due to the new rule changes and point system, we decided that a goalie robot is best for us to pick and go with. We came up with all sorts of ideas such as: * Expanding ramps on all sides * Nets on top to "steal" balls * Robot deterrents such as wheels/wipers * Expand "up" and "down" * active ramps with pneumatic pistons * moving laterally * 24V the system * a hook for the metal bar at the goal Mind you, Gridlock is perhaps the longest and hardest to plan robot we had for Tagisang. We designed gridlock for about a month or less. We used Autodesk Inventor 2014 to create the 3D model. Robot assembly2.jpg Robot assembly3.jpg Robot assembly.jpg Robot.png Gridlock's bill of materials : # a lot of aluminum l beams, flat beams, and about 18 ft of aluminum C-channels # 4 Toyota Landcruiser surplus wiper motors; 2 for locomotion, 2 for the robot deterrents # 1 dismantled 18v cordless Ozito Drill for the release mechanism of the expanding mechanism # 1 large brass door hook as the lock of the expanding mechanism # 1 5.5in capacity gas spring (though we bought 2) # a lot of piano hinges for the flaps # netting # felt for padding # GI sheets for the ramps # rubber tourniquet (medical rubber bands used to constrict blood flow) # 2 rubberized cart wheels, about 6 in diameter # door springs and pvc pipes for them to slide in # 2 12v batteries # Arduino and TRC Vehicleboard; 2 motor drivers # assorted wires # myriads of cable ties The Build Gridlock took about a month to make. The biggest setbacks were mechanical/geometric problems and figuring out how to make it all work. With so many moving parts, Gridlock was almost bound for these problems. We even had to trim the length of the robot because we underestimated the additional length overlapping aluminum and padding would have. Initially we just cut out 0.25 in on both sides, but perhaps for the best we should have allowed 1 in or even more total allowance. We weren't able to simulate the moving parts of Gridlock in the CAD model, just the basic structure of it, so we basically created the mechanisms and means of movement, expansion, and installation on the fly. The Expansion System The expansion system was the first problem. Making it expand top and bottom in such a way that it would satisfy the "interpretation of what judges think of the rules" felt like making the mechanism almost illogical to do. So what we devised for Gridlock is that the main chassis with wheels "goes down", in effect being longer downward. The thing is, moving something upward can be said of the same as well. However, we chose the former method so that AT LEAST we can say the robot extended DOWNWARD, if that isn't enough for the judges to accept. Thus the robot has a "umbrella" on top which moves upward as the aluminum "stand" under the robot goes up as well. Inside the Aluminum square pole and is the hidden upper portion of the AL stand however, is 2 long Al L beams that align to each other like a square tube with a bent sealed top. Its function is to allow the the whole upper "umbrella" system to hoist up the AL stand with a screw that goes through the AL square pole and through the 2 AL L beams - with a sealed top but with a gap between each end of the L beams. The Al square pole is guided by 2 shorter C channels formed into a bigger square pole. We underestimated the strength of the gas springs and the dual gas spring setup is abandoned. It was strong enough to bend aluminum and damage the hook release. Kuya then placed the gas spring in the middle instead. Looking for a suitable release mechanism is also one of our big problems. We brainstormed and tried several methods from the hook to a door lock style mechanism to a seatbelt-like quick release to tape then finally back to the hook. The drill motor did wonders and easily pulls the hook of the strained expanding system. we used ribbon as the line. The upper and side flaps are hinged with piano hinges and actuated by door springs, inside pvc pipe to guide its movement. The Locomotion Coupling the wheels to the motors is another huge problem. At first, we had the wheels and motor shafts threaded in a machine shop and applied threadlock onto the threads. The machine shop gave us its own setbacks as well. We grossly overestimated threadlock, because it could barely keep the wheels on the shaft even after overnight curing. We opted to have it welded on instead by Kuya, though that would mean it will be a permanent fixture and risk destroying the motors and even the wheels since it'd be stuck to the destroyed motor. We also had big problem with shorting wiper motors because the body of the wiper motors are made to be the ground of the motors. What we had to do is to isolate the motors from the aluminum chassis with duct tape. Throughout our build time, we ran the motors in 24v, which proved to be the culprit of our robot's odd driving. The Practice We also made a simulation of the goal and shooting conditions at the contest for our training. It helped us see what are the holes in the idea and see for ourselves how everything would work, including the likeliness of human shooters scoring. We let our driver also practice this way but due to unending problems to our constantly changing robot, he wasn't able to practice well. IMG 20140820 192713.jpg|Practice Goal setup Robot Review, Pre-Contest IMG 20141023 201350.jpg IMG 20141015 104359.jpg IMG 20141011 184457.jpg IMG 20141007 181809.jpg The Contest Proper IMG 20141030 094547.jpg Overlooktrc2014.jpg Gridvsbnhs.jpg Visit the Tagisang fb page for more pictures: https://www.facebook.com/tagisangrobotics/photos_stream Plenty of adjustments at contest. The robot was basically redone there. We scrapped the hook because it caused more problems and is rather hard to use. The height of the bar was different compared to the one we thought it was. We had to make a release aid for the diagonal flaps to fall out properly, otherwise it would fold inward and cause the side ramps to stay lifted, which were also proved to be a distraction to the driver because he had to wiggle them out to fall out properly. The biggest blow was forcing the 24v system to work. We thought that initially the caster wheels were the cause and ended up going round the bush replacing it with so many alternatives. It cost us many opportunities to practice controlling the robot. Perhaps we should've learned to let go some features that simply didnt work in the given time. Knowing when to give it up though is still very hard, because even the expanding top/bottom problem was like that. =Epilogue= Special thanks to Kuya Ed Rubio, who practically made the robot himself, with our awkward attempts to guide him. Special thanks to our coaches, janitors, alumni, and everyone else who had made this possible. Overall, We feel that we were able to showcase gridlock's true worth, but not really its full potential. There are many lessons to be gleaned from this journey, but nevertheless, it has been a wonderful and enjoyable experience. The journey has been against all odds, and throughout it all, we can undeniably say that everything was under control by God. To God be the glory! IMG 20141030 171654.jpg Tagisang2014wgridlock.jpg Category:Tagisang Category:Competitions Category:Local Competitions